Oil Production from the Bakken Formation: A Short History By Julie LeFever

Once again the Bakken Formation is in the news as oil Plays of the Bakken Formation prices rise. Over the years the State of North Dakota has Pre-1987 seen several episodes of drilling to try to recover a portion of the immense amount of oil that this formation has generated. Early exploration for Bakken oil was by conventional Beginning with the initial discovery in 1953, the Bakken has vertical drilling and began with the discovery of Antelope been a significant contributor to the North Dakota economy. Field (Fig. 1B). Oil was recovered in Antelope Field from the Over 44 million barrels of oil have been produced from this Bakken Formation in 1953 when an unsuccessful Madison formation, and there is a large potential to add to that total test (Stanolind Oil – #1 Woodrow Starr (SWSE Sec. 21, with the current oil play. T152N, R94W)) recovered oil on a drill stem test. The well was completed for 536 barrels of oil (BO) and essentially no Although a significant quantity of oil has been produced water. Offsets to this well were disappointing, so further from the Bakken it is only a fraction of what the Bakken is development did not occur until 1956 when Northern Pump capable of producing. Long considered to be a premier source Co. drilled and completed the #1 Ella Many Ribs for 320 rock, science places the oil potential of the Bakken between barrels of oil per day (BOPD). The majority of the wells for 10 and 400 billion barrels of oil (L.C. Price, pers. commun.). the field were drilled in the 1950’s and 60’s. During the A comparison of the past drilling activity in the Bakken 1980’s, renewed interest in the Antelope structure resulted Formation with the current play provides a new insight into in the drilling and completion of a number of wells along the how technology can potentially unlock oil from this virtually southern end of the feature. untapped resource. Wells were completed in Antelope Field in both the Bakken Formation and underlying thin, highly burrowed porous silty sandstone referred to as the “Sanish Sand”. Early wells in the field were drilled with a salt-based mud system; this was later changed to an oil-based mud system to prevent dissolution and the resulting casing collapse of the overlying Madison salt section. Typical completion methods consisted of perforating the well followed by a sand-oil fracture stimulation treatment. The prominent occurrence of natural fractures along the structure enhanced the production.

The field has been highly successful. Fifty-two wells produce or have produced along the structure with an average initial production (IP) of 209 BO. The field produced 67,798 BO in 2003 and has a cumulative total in excess of 12.5 million barrels of oil and 10 billion cubic feet (BCF) of gas.

After the development of Antelope Field, exploration of the Bakken proceeded slowly. The tight, impermeable reservoir of the Bakken required

Figure 1. – A) Location map of the Williston basin. B) Fields with production from the Bakken Formation along the Nesson anticline. C) Fields producing from the Bakken along the “Bakken Fairway”. D) North Dakota fields with Bakken production adjacent to Richland County, Montana.

Vol. 32, No. 1 NDGS Newsletter Page 1 natural fractures in order to produce. The next discovery in Successful wells were also drilled with oil-based fluids. the formation did not occur until 1961 when Shell Oil The Bakken shale is an oil-wet reservoir and wells that had Company used seismic methods to discover Elkhorn Ranch water introduced to the system had difficulty unloading the Field along the depositional limit in what is now known as water. Introduction of water into the system could be through the “Bakken Fairway” (Fig. 1C). drilling fluids, acidizing, or fracture stimulation methods. Acidizing Bakken wells also creates additional problems, as This was a typical Williston basin discovery. The well the treatment may react with the pyrite present in the shales was an Ordovician Red River test that recovered oil on a drill forming an iron hydroxide precipitate. stem test in the Bakken. The Shell Oil Company - #41X-5-1 Government (NENE Sec. 5, T143N, R101W) was The long history of vertically drilling the Bakken set the subsequently completed and tested 136 BOPD and a trace stage for horizontal drilling. Vertical drilling established that of water. However, the well was located in a remote area the wells with significant vertical production were draining and was often shut-in during inclement weather. It produced fractured reservoirs. It also helped to determine the proper an average of 50 BOPD until 1964, when its casing collapsed methods necessary to successfully complete a Bakken well. and the well was plugged and abandoned. Further The production decline curves display a characteristic rapid development of the Bakken did not occur in this area until drop in production over the first year, followed by steady 1976. production with virtually no decline.

Another notable discovery at this time was along the Post-1987 Nesson anticline. The #1 B.E. Hove (SWNW Sec. 2, T154N, R95W) was drilled by Pan American Petroleum Corporation Drilling activity along the “Bakken Fairway” changed in Hofflund Field. During its initial test, the well flowed 756 greatly after Meridian Oil, Inc. drilled and completed the first BOPD and 3 barrels of water per day (BWPD). This well horizontal well in the Bakken Formation. The #33-11 MOI, was perforated and then treated with 10,000 gallons of acid. was initially drilled vertically. It was cored, logged, and drill Production problems resulted in two workovers with stem tested, all of which indicated that the formation was additional acid treatments followed by a water-based fracture tight. Meridian then backed up the hole and kicked off at stimulation. The well produced for two years with a 9,782 ft. Horizontal drilling was attained at 10,737 ft cumulative production of 62,700 BO. It was plugged in 1969 (measured depth) with a resulting radius of 630 ft. The well due to non-commercial production and collapsed casing. was completed on September 25, 1987 for 258 BOPD and 299 thousand cubic feet (MCF) of gas. The well had a Continued exploration for Bakken oil occurred along horizontal displacement of 2,603 ft and is now producing in the “Bakken Fairway”. Wells along the “Bakken Fairway” the upper Bakken shale that is 8 ft thick. The decline curve are not limited to Bakken production. Most of the fields have for the #33-11 was remarkably stable for the first two years multiple pays that are associated with structural features until additional nearby wells came online. The well has such as folds, faults, or both. These features contribute to produced 357,671 BO and 6,381 barrels of water (BW) the fracturing of the Bakken and may enhance the quality of through December 2003. the reservoir. The ability to produce from other formations makes it much more cost effective. The drilling activity in The success of this well set off the previous notable play this area began in the late 1970’s. dealing with the upper Bakken shale. Operators were eager to use horizontal well technology to encounter more fractures In the “Bakken Fairway”, the formation exhibits an on- and thus produce more Bakken oil. As in any play, there was lapping relationship. The stratigraphy of this area is relatively a learning curve. The #33-11 MOI took 57 days to drill and simple. The Three Forks Formation is overlain by the complete; 27 days to drill the vertical borehole and 12 days remaining, stratigraphically higher, members of the Bakken. to drill the horizontal section at a cost of $2 million. The The Bakken, in turn, is overlain by the carbonates of the third set of ten wells drilled by Meridian, further down the Lodgepole Formation. As the Bakken thins, it becomes much learning curve, had an average cost of $1.08 million and took more susceptible to fracturing and, as shown by previous 35 days to drill. By the end of the play, Meridian Oil, Inc. was drilling, fracturing is necessary for production. touting the fact that they could drill and complete a horizontal Bakken well for essentially the same price as the drilling and Successful wells were those that encountered natural completion of a vertical well, around $900,000. Successful fractures. Less than successful wells, those not encountering wells were capable of producing high volumes of oil (IP’s in natural fractures, were perforated in the Bakken shale and excess of 1900 BOPD). produced for a while. Oil was collected or diesel was substituted for produced oil and used as part of a sand/oil The play continued into the early 90’s. During the play, fracture stimulation treatment. This treatment was necessary 22 operators drilled horizontal wells. Half of these operators for economic reserves. were involved in one or two well programs. The early horizontal wells were medium-radius wells with one 2500

Page 2 NDGS Newsletter Vol. 32, No. 1 to 3500 ft horizontal leg. They were drilled underbalanced successively higher member within the Bakken has a greater or slightly underbalanced with an inverted mud system and distribution than the preceding member. The members completed with slotted liners tied back to casing. It was not eventually pinch out against the Three Forks Formation in uncommon to be stuck in the hole extending the drilling time southern Richland County. In the main play area, only the and the expense of the hole. The wellbore still had the same upper two-thirds of lithofacies 2 and the overlying upper sensitivity factors to fluids that were encountered in the shale member of the Bakken remain. Impermeable rocks of vertical holes. If the wells did not encounter naturally the underlying Three Forks and the overlying Mississippian occurring fractures there was no production and no ability Lodgepole as well as the pinch out of the facies provide the to stimulate the wellbore in an attempt to connect with a trapping mechanism. regional fracture system. The Balcron Oil - #44-24 Vaira (SESE Sec. 24, T24N, Another problem occurred with the horizontal Bakken R54W) has a well defined log characteristic for the porosity shale play. Interference between wellbores was well zone of interest (Fig. 3). The distinctly cleaner gamma-ray documented to extend over a large portion of the Fairway. log is typical of the pay zone and is mappable over a large Horizontal wellbores could have an effect on a vertical area (Fig. 4). A log porosity of approximately 10% can be wellbore 2500 to 3000 ft away. The large scale regional calculated for the pay zone. fractures are highly interconnected, causing pressure and fluid communication to extend over several sections. The Three Forks consists of interbedded tan fine-grained Interference between wells would not show up immediately, sandstone to siltstone and light brown claystones. Structures but would generally become apparent after a month of include parallel laminations that are locally brecciated continuous production. Routine maintenance included (probably due to the dewatering of the sediments) and ripple periodic removal of salts and fines that would accumulate in laminations. The sequence is highly cemented and the bottom of the wellbore. impermeable, forming a good bottom seal.

2001 to Present

The Bakken Formation is once again in the spotlight. The current drilling activity began in 2001 and is largely centered in Richland County, Montana (Fig. 1D). Success is pushing the play to move into North Dakota. Leasing activity within the State has been high over the past several months in anticipation of the play moving across the border. To date, there is one reportable well drilled with numerous permits for additional wells.

Middle Member Bakken Play Richland County, MT

The stratigraphy of the play area is relatively simple Figure 3. – Typical wireline log from Richland County, Montana. (Fig. 2). It consists of the Bakken Formation onlapping the Log shows the pay section that is the horizontal target. underlying Devonian Three Forks Formation. Each

Figure 2. – Stratigraphy of the Bakken Formation in the play area Figure 4. – Isopachous map of the pay zone within Richland of Richland County, Montana. County, Montana (CI=2 ft).

Vol. 32, No. 1 NDGS Newsletter Page 3 The lower Bakken shale is absent over most of the play terminates the trend. The lithofacies is traceable in North area in Montana, and the middle member sits unconformably Dakota but shifts location due to the presence of the salt. on the underlying Three Forks. It is common to see abundant debris along the contact between the Three Forks and the North Dakota middle member siltstone. The basal portion of the Bakken middle member consists of a brownish-grey, argillaceous Production from the middle member of the Bakken is siltstone. Burrowing and small clay drapes are present. not new to North Dakota. There have been a few wells, Immediately above the clay drape section is a dolomite or an primarily along the northern Nesson anticline that have been equivalent bioturbated siltstone/sandstone sequence that is perforated in and produced from the Bakken middle member the producing zone in Montana. This is then overlain by a (Fig. 1B). The Bakken in these wells was not the primary thin transitional zone and the upper Bakken shale. target; that was deeper usually the Devonian Winnipegosis Formation. Wells producing from the Bakken were The middle member of the Bakken Formation gradually perforated because they were reaching their economic limit thickens towards the center of the basin in North Dakota or had no other production potential. Production from these where it attains a maximum thickness of 90 ft (Fig. 5). An wells was limited. isopachous map of pay section in lithofacies 2 shows why the play is where it is (Fig. 4). There is a noticeable trend Five lithofacies within the middle member of the Bakken approximately 2 to 3 townships wide by 72 miles long. Along Formation have been identified throughout Montana, North this trend, the middle member reaches a thickness of 35 feet Dakota, Saskatchewan and Manitoba (Fig. 6). In ascending and then thins somewhat before continuing its gradual order, they are: thickening towards the center of the basin. The thickness of 1) Lithofacies 1 consists of light grey, greenish-grey, the producing interval reaches a maximum of 14 feet and has or brownish-grey argillaceous siltstone. It is generally consistent log porosity between 8 and 12%. massive, cemented with calcite, and has scattered pyrite nodules and fossils (crinoids and brachiopods). Locally the The other porosity zone contoured (to the north) is one unit is burrowed. Porosity is intergranular. that occurs within the middle member (Fig. 4). This zone 2) Lithofacies 2 consists of greenish-grey to brownish- has limited distribution and occurs between the two shales. grey, argillaceous siltstone or sandy siltstone to brownish- Facies changes, like this, within the middle member suggest grey, very fine grained sandstone. Small scale clay drapes are that these may be common and potentially productive. Other present, as well as burrows. additional lithofacies identified in other areas of the basin 3) There are three parts to Lithofacies 3. The upper can be easily mapped and are potentially productive (LeFever and lower third consists of wavy to flaser bedded, light to et al., 1991). medium grey, argillaceous to sandy siltstones and brownish- grey, very fine-grained sandstones with local claystones. The The other notable feature of the play is the influence of middle third consists of a medium grey, dark grey, or greyish- the Devonian Prairie Salt. Indicated by the solid black line on tan, fine- to medium-grained sandstone that may be massive, Figure 5, the trend present in Montana appears to be directly cross-bedded, or thinly laminated. related to the dissolution of the underlying salt. The trend 4) Lithofacies 4 consists of alternating medium grey that is strongly present in Montana can be traced into North argillaceous siltstones, light to medium grey very fine-grained Dakota as far as Mondak field. As the state line is crossed, sandstones and dark grey shale. The unit is thinly laminated, the edge of the Prairie shifts to the south and locally displays planar and ripple laminations, is moderately bioturbated in places, and locally cemented. 5) Lithofacies 5 is a medium to light grey, massive to wispy laminated siltstone that is generally cemented.

Most of the production from the middle member in North Dakota prior to the start of this oil play was from Lithofacies 3. As described above, it is a sandstone-siltstone bed that occurs throughout the northern portion of North Dakota and extends into Saskatchewan and Manitoba. It is highly productive in Saskatchewan with a long production history from Rocanville and Roncott Fields. In North Dakota, production from this facies is marginal and found in 9 fields along the northern Nesson Anticline including Temple, Figure 5. – Isopachous map of the middle member of the Bakken Stoneview and North Tioga (Fig. 1B). It has an easily Formation (CI=5 ft). Richland County, Montana and McKenzie mappable wireline log characteristic throughout North County, North Dakota are shaded. Heavy black line represents Dakota with a notably cleaner response on the gamma-ray the edge of the Devonian Prairie Salt. curve (Fig. 7).

Page 4 NDGS Newsletter Vol. 32, No. 1 All of the wells producing from Lithofacies 3, with the Drilling and Completion of Bakken Horizontal Wells exception of one, were completed as a bailout zone. This facies is generally less argillaceous, coarser grained and better Technology has finally caught up to the Bakken cemented than the rest of the middle member. Thickness of Formation. The ability to fracture stimulate these horizontal the pay zone is 12 ft with 7 to 12% porosity. Well wells is what makes this play work. Whereas, in the late completions generally consisted of acidization and a gelled 80’s-early 90’s wells had to rely on encountering natural water-sand fracture treatment. It is interesting to note that fractures to supply the oil; the wells in this play create their the best producer in Stoneview underwent the least amount own fractures. of acid treatment. Wells generally consist of two 4000 to 5000 ft laterals drilled on a 1280-acre spacing unit. Also, unlike the previous play, the middle member is drilled with saturated brine instead of inverted mud. The zone generally has between 7 to 12% porosity, a permeability of .01 to .02 md, and 70 to 80% oil saturation. Once drilled, the well is then treated with a 650,000 to 1 million pound gelled water-sand frac. Due to high gas content, it is common for the wells to flow for quite a while before having to be put on pump.

The per well cost is approximate $2.2 million with the potential for the well to produce 500 to 700 BOPD initially, leveling off at 250 BOPD with virtually no water.

North Dakota

The first horizontal well of this play in North Dakota was drilled and completed on March 3, 2004 in Lithofacies 3. It was a re-entry of a 1981 Madison test in Sadler Field by Joilette Oil, Inc. - #1-17R Robert Heuer (SESE Sec. 17, T161N, R95W). One lateral leg was drilled with a sidetrack. The lateral was then treated with gelled water-sand frac and tested. Reported results for the initial potential test were 87 Figure 6. – Diagram showing the lithofacies present in the middle BOPD, 150 MCF of gas, and 142 BWPD. Cumulative member of the Bakken Formation throughout the Williston basin. production to date on this well is 14,838 BO, 12,021 MCF Lithofacies 2 is productive in Richland County, Montana. of gas and 4,515 BW. Lithofacies 2 and 3 are productive in North Dakota. Although results were not like what has been seen in Montana, this is a different facies, which will probably require adjustments to the completion method. This is also the initial well in the play and there is always a learning curve.

Conclusions

The middle member Bakken play is well on its way in Montana. Prior to the start of the play, Richland County had 80 producing fields in the county with 23 producing from the Bakken Formation. Data from the Montana Board of Oil & Gas show that 50 wells produced from the Bakken prior to the current play. By the end of August 2004, 127 wells had production from the Bakken. Ninety-four horizontal wells have been drilled into the Bakken with 92 producers, one PNA (plugged and abandoned), and one dry hole. Additionally, numerous permits have been issued. The impact Figure 7. – Typical log of the producing section along the northern of the current activity is easy to see based on the total Nesson anticline. Lithofacies 3 is the pay section of the Joilette production from Richland County. Production has doubled Oil (USA), LLC. - #1-17R Robert Heuer. since 2003, and is still increasing.

Vol. 32, No. 1 NDGS Newsletter Page 5 The same facies that produce in Montana are present The ability to stimulate horizontal wells is driving the and potentially productive in North Dakota. There may be play. Past history has shown that the Bakken doesn’t give up slight changes in composition, locally rocks may be slightly its oil easily. It requires some sort of fracture system and in more silty/sandy or more argillaceous so adjustments may the past nature controlled the system. Now, that fracture have to be made to drilling or completion techniques. system can be artificially generated to open the Bakken up to production. The drilling of these wells with 2 long horizontal North Dakota has a thicker section of the Bakken middle legs and stimulating them with a large frac treatment is member. Additional section may suggest additional targets expensive ($2.2 million) but can lead to a very profitable well for horizontal drilling. Production has already been proven especially at current oil prices. The potential for success from one of the higher lithofacies. There is also an additional from horizontally drilling the middle member is great. There section at the base of the Bakken, the “Sanish Sand”. This is will be a learning curve associated with these wells, just as a significant producer at Antelope Field and occurs throughout there is in every other play. the “Bakken Fairway”. This highly burrowed sandstone/ siltstone facies is probably related to salt collapse and may References Cited be productive in the fairway area. Another potential target LeFever, J.A., Martiniuk, C.D., Dancsok, E.F.R., and Mahnic, may be the lower Lodgepole Limestone between the upper P.A., 1991, Petroleum potential of the Middle Member, Bakken shale and the “False Bakken”. Detailed mapping of Bakken Formation, Williston Basin, in Christopher, J.E. all the zones will be required to determine the best location and Haidl, F.M., eds., Sixth International Williston Basin to tap into the Bakken resources. Symposium: Saskatchewan Geological Society Special Publication No. 11, p. 74-94.

Page 6 NDGS Newsletter Vol. 32, No. 1